key: cord-1046920-obv65dzy authors: Tahir Huyut, Mehmet; Ilkbahar, Fatih title: The Effectiveness of Blood Routine Parameters and Some Biomarkers as a Potential Diagnostic Tool in the Diagnosis and Prognosis of Covid-19 Disease date: 2021-06-02 journal: Int Immunopharmacol DOI: 10.1016/j.intimp.2021.107838 sha: 1dac145d36f764767165533e66af9ac057671360 doc_id: 1046920 cord_uid: obv65dzy Since February-2020, the world has been battling a tragic public-health crisis with the emergence and spread of 2019-nCoV. Due to the lack of information about the pathogenesis-specific treatment of Covid-19, early diagnosis and timely treatment are important. However, there is still a lack of information about routine-blood-parameteres (RBP) findings and effects in the disease process. Although the literature includes various interventions, existing studies need to be generalized and their reliability improved. In this study, the efficacy of routine blood values used in the diagnosis and prognosis of Covid-19 and independent biomarkers obtained from them were evaluated retrospectively in a large patient group. Low lymphocyte (LYM) and white-blood-cell (WBC), high CRP and Ferritin were effective in the diagnosis of the disease. The d-CWL= CRP WBC ∗ L Y M and d-CFL= CRP ∗ F e r r i t i n LYM biomarkers derived from them were the most important risk factors in diagnosing the disease and were more successful than direct RBP values. High d-CWL and d-CFL values ​​largely confirmed the Covid-19 diagnosis. The most effective RBP in the prognosis of the disease was CRP. (d-CIT) = CRP*INR*Troponin; (d-CT) = CRP*Troponin; (d-PPT) = PT*Troponin*Procalcitonin biomarkers were found to be more successful than direct RBP values and biomarkers used in previous studies in the prognosis of the disease. Finally, an open-access data source consisting of RBC was created for studies to be carried out in the fight against COVİD-19. In this study, biomarkers derived from RBP were found to be more successful in both diagnosis and prognosis of Covid-19 than previously used direct RBP and biomarkers. 2019-nCoV, which is a member of the Coronavirial and Nidovirale family, is an enveloped, positively oriented, single-stranded RNA beta coronavirus [1] . The virus originated in bats and was transmitted to humans through animals in Wuhan, China in December 2019 and spread from there all over the world. The Covid-19 outbreak (caused by severe acquired respiratory syndrome-coronavirus-2 [SARS-CoV-2]) posed and continues to pose a major threat to public health globally [2, 3] . Symptoms were generally seen as fever, cough, sore throat, shortness of breath, fatigue as well as weakness [1] . Moderate and general patients have milder symptoms and a good prognosis, but severe and critically ill patients are difficult to treat and have a high mortality rate [4, 5] . Unfortunately, information on early prediction factors for disease progression is relatively limited. Due to the lack of constructive information about the pathogenesis and specific treatment of Covid-19, early diagnosis and timely treatment are of clinical importance [6, 7] . The most common laboratory findings, including hematological and biochemical parameters, play an important role in the initial screening for Covid-19 [6, 8] . Routine laboratory parameters have great clinical importance in predicting the diagnosis and progression of the disease [5] . In addition, while the number of Covid-19 cases is increasing day by day, there is limited information about the hematological and laboratory findings of the disease and their effects on the disease [4, 5, 9] . For this reason, many academic studies are conducted to determine the change of blood routine laboratory parameters in Covid-19, prognosis, morbidity, mortality and their relationship with other comorbidities. Although the literature includes various attempts to address such issues, the place and importance of routine blood laboratory parameters in Covid-19 disease is not fully understood [4, 5] . Although the clinical features of Covid-19 have been identified, large sample size studies representing laboratory abnormalities of these patients are still needed [7, 10] . Therefore, the relationship between Covid-19 disease and laboratory parameters should be supported by large data sets. In addition, more successful biomarkers continue to be used as alternatives to current clinical approaches in the diagnosis and prognosis of the disease. Within the scope of these important goals, this study compared the effect of predictive routine blood parameters frequently used in the literature in the prediction of the diagnosis and prognosis (in the distinction of subjects in all services unit and intensive care unit) and the success of independent biomarkers derived from these blood values in this sense. In addition, the success of the derived biomarkers and the biomarkers previously used for this purpose in the literature were compared. With the results to be obtained, it is thought that this study will guide useful strategies for clinicians in predicting the diagnosis and prognosis of the disease. Finally, we have created our open access data source for studies to be conducted in the detection of Covid-19 disease, its prognosis and the relationship structures between routine laboratory parameters (http://covdats.duzce.edu.tr). This retrospective single-center study was conducted in accordance with the 1989 Helsinki Declaration. This study was approved by the Ministry of Health and the Ethics Committee. Data matching our criteria were collected from Erzincan Binali Yıldırım University Mengücek Gazi Training and Research Hospital information system between March 2020 and December 2020 and were included in the study. The research only covered people over 18 years old. Laboratory data of the patients were the first blood values measured at the time of first admission to the hospital. With the diagnosis of "Covid-19 identified virus", patient information was scanned and the information of a total of 2648 patients was reported. The diagnosis of Covid-19 was defined only in cases detected as SARS-CoV-2 by rRT-PCR in nasopharyngeal or oropharyngeal swabs at the dates covered by this study in our hospital. Routine laboratory parameters generally used in the prognosis of Covid-19 were obtained from all patients. Biomarkers frequently used in the diagnosis, prognosis and mortality estimation of the disease were calculated for all patients. When the data were being scanned, they were reported in two groups, those treated in the intensive care unit (ICU) and subjects in all services unit (non-ICU). In addition, an equal number of control groups (2648 individuals) with the patient group were determined. The Sysmex XN-1000 Hematology System (Sysmex Corporation, Kobe, Japan) was used to perform cell blood count of individuals in the experimental and control groups. Biochemical tests were analyzed from serum by spectrophotometric method using Beckman Coulter Olympus AU2700 Plus Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). Ferritin was evaluated by a chemiluminescence immunoassay (Centaur XP, Siemens Healthcare, Germany). Prothrombin time (PT), activated partial prothrombin time (aPTT), and fibrinogen were determined with a fully digital coagulation device of Ceveron-Alpha (Diapharma Group Inc., West Chester, Canada). C-reactive protein (CRP) was measured by the nephelometric method on the BN ™ II System (Siemens, Munich, Germany). Procalcitonin (PCT), D-dimer and Troponin were analyzed from whole blood on the AQT90 flex Radiometer® (Bronshoj, Denmark). Biomarkers from previous studies 4-5 used in predicting the prognosis and mortality of Covid-19 disease were identified and summarized below. The effect of these biomarkers in the diagnosis of Covid-19 disease and in the separation of inpatients into intensive care / service (all non-intensive care units) units (ie the prognosis of the disease) was investigated. Platelet count, CRP: C-reactive protein count. Biomarkers have been derived for the diagnosis of Covid-19 disease and estimation of the separation of inpatients into units (intensive care/all non-intensive care units) (i.e., for the prognosis of the disease). In the conclusion part, the performances of the biomarkers were compared. The new biomarkers that have been derived are summarized below. Independent biomarkers derived from our approach in this study will provide an alternative to clinical approaches in predicting the diagnosis and prognosis of Covid-19. While deriving biomarkers, our research question was: Will the biomarkers we derive be more successful in predicting the diagnosis or prognosis of the disease than existing biomarkers in the literature or direct blood values? In order to determine the predictors that are effective in the diagnosis of the disease, the difference of the predictors between the ICU-Control and nonICU-Control groups was examined. In addition, in determining the predictors that are effective in the prognosis of the disease, the difference of predictors between non-ICU and ICU was evaluated. Direct values of blood values were used when designing biomarkers to predict Covid-19 or to separate patients into non-ICU and ICU units (ie, predict the prognosis of the disease). For this purpose, the following criteria were applied in the reporting of many different mathematical ratios (independent biomarkers) tested on the direct data of blood values: As a result of the Multivariate Logistic Regression, which was run to estimate the individual effect of the predictors for the diagnosis and prognosis of the disease, the predictor was 1) significant odds-ratio (OR) (ie, have significant influence); 2) significant AUC ( area under the receiver operation characteristic curve (ROC)) value. The dependent variable for the diagnosis of the disease was patient-control (binary), the dependent variable for prognosis was non-ICU and ICU (binary). In addition, if the predictors were evaluated together, would there be an increase in success in predicting the diagnosis and prognosis according to their individual evaluations? Therefore, decision trees suitable for the use of readers-clinicians were drawn for the collective evaluation of predictors that were found to be effective and successful in the diagnosis and prognosis of the disease. CHAID algorithm was used in drawing decision trees. The values of the predictors in terms of accuracy, sensitivity, specificity, area under the ROC curve (AUC) were obtained. Since false negatives for diagnosis (i.e., Covid-19 positive patients who were classified as negative and possibly allowed to go home) were more harmful than false positives in this screening task, accuracy (success in identifying patients and healthy individuals in total) and AUC (how positive were those determined as positive actually were) ) as the main quality criteria. Similarly, in differentiating patients between non-ICU and ICU units (i.e., in determining the prognosis of the disease) false negatives (i.e., covidites in non-ICU while they should be in the ICU) are more harmful than false positives in this screening task, was taken accuracy and AUC quality criteria evaluated as. CHAID analysis is a nonparametric method that can model categorical and continuous variables together, providing reliable estimates in large samples. In addition to these advantages, it can detail the relationships between independent variables and provide easy-tounderstand outputs in the form of trees, even in the most complex models. CHAID analysis with these advantages has a wide range of uses. Contol group on admission Significant difference was observed between the non-ICU and ICU groups in all biomarkers (derived and available biomarkers) except d-TI. Here Table 1 3 Contol group on admission biomarkers (derived and available biomarkers) were higher in ICU patients. Here Table 2 3 Here Table 3 3 Those who were treated in the hospital with the definitive diagnosis of Covid-19 (non-ICU and ICU total 2648 people) were determined as a positive group. No Covid-19 patients (2648 people who applied to the hospital for other complaints) were determined as the control group (negative group). We obtained the crude odds-ratio (OR) after performing multivariate logistic regression analysis to identify predictors that may affect the diagnosis of Covid-19 ( Table 4 ). Here Table 4 3. 5 Those who were treated with the definitive diagnosis of Covid-19 were determined as a positive group. The control group that was not diagnosed with Covid-19 was determined as a negative group. The performances of the predictors with significant odds-ratio in the diagnosis of the disease (Table 4) were obtained (Table 5 ) and cut off values were determined by drawing the ROC curves (Fig 1) . When looking at the diagnostic values of blood parameters and independent biomarkers in Table 5 , it is seen that the specificty value in favor of sensitivity sometimes decreases. Since false negatives for diagnosis were more harmful than false positives in this screening task, was taken accuracy (ACC) and AUC as the main quality criteria (see section 2. 6). When Table 5 ) (the all p values < 0.001). Here Table 5 3. 6 Table 6 ). Here Table 6 3. 7 Covid-19 patients treated in the ICU unit were determined as a positive group, non-ICU were identified as negative group. The receiver operating characteristic (ROC) curve was created to analyze the diagnostic performance of various routine blood parameters and biomarkers (parameters in Table 6 ) whose effectiveness in determining the diagnosis of ICU (significant OR) on admission (Fig 2) . When we look at the diagnostic performance values of blood parameters and independent biomarkers in determining ICU patients in Table 7 , it is seen that the specificity value in favor However, biomarkers derived in this study were found to be more successful in identifying ICU patients (true positives-AUC) and identifying non-ICU and ICU patients (ACC) than older biomarkers and routine blood values (see Table 7 ). When looking at the AUC and the Here infection are protected from a second infection [13, 14, 15] . With these clinical features of Covid-19, there is no specific treatment approved yet [13] . Therefore, predicting the diagnosis and prognosis of the disease is important for early intervention [4]. In a sense, it is possible to determine both the diagnosis and the prognosis of the disease with routine blood values (biomarkers) that are easier to access, more economical and faster. Similarly, one study said that dynamic surveillance of the Peripheral blood system, particularly eosinophils, helped predict severe cases of Covid -19 [5] . Indeed, this study found abnormal peripheral routine blood results in Covid-19 patients. These advantages have further increased the importance of routine blood values and will be a predictive aid in the current crisis. In the study conducted for this purpose, changes were found in most of the routine laboratory tests and biomarkers. It was observed that these changes were more in intensive care patients. This study showed that older age and male sex are seriously threatened in terms of both 16, 17, 18, 19] . In this study, the efficacy of routine blood values and biomarkers in the diagnosis of the disease (patient-healthy) and classification of patients into non-ICU or ICU (prognosis of the disease) were evaluated. In this study, a significant increase was observed in routine blood parameters of CRP, D- [5, 23] . Similarly, previous studies reported that lymphopenia is common in Covid-19 cases and is an important and reliable indicator of disease severity [5, 24, 25] . It was stated that lymphopenia can be caused by the Covid-19 virus, either directly (for example, being the target of the virus with the ACE2 receptors on them) or indirectly, which suppresses lymphocyte production or shortens the half-life of lymphocytes [26, 27, 28, 29] . In this study, biomarkers (derived) that are effective in the diagnosis of the disease were found to be more successful than the diagnostic performance of direct blood values (Table 5) . Specifically, d-CWL and d-CFL were the highest in the diagnosis of the disease, AUC: 77.8 %, ACC: 77.6% and AUC: 78.3%, ACC: 79.54%, respectively (Fig 1) . LYM, NLR, LCR, PLR, d-NLR, which were used as important biomarkers in the diagnosis of the disease in previous studies [4, 5, 19, 20, 29, 30, 31] , performed lower than the d-CFL and d-CWL produced in this study (Table 5 ). In addition, it was observed that biomarkers derived to determine the diagnosis of the disease were not affected by gender and age, while NLR, which is frequently used in the diagnosis of the disease in the literature, was found to be affected (adjusted odds-ratio, see Table 4 ). Accordingly, the d-CWL and d-CFL biomarkers that we have derived instead of direct blood values and existing biomarkers can be used in the diagnosis of the disease. However, these diagnostic achievements of blood values and biomarkers are individual. For a user-friendly guide that can be used in suspicious cases in the diagnosis of the disease, a decision tree that evaluates the predictors together was drawn (Fig 3) . The total classification accuracy rate of the decision tree, which was created in a statistically significant way with eight predictors in the diagnosis of the disease, was 81.6 %, the accuracy rate for classifying patients was 85.2 %, and the accuracy rate for classifying the healthy was 77.9 %, and it was promising ( Table 8 ). The presence of CRP and its derived biomarkers at the root of the tree was an interpretable confirmation of the accuracy of our Decision Tree (Fig 3) in the approach to diagnosing the disease. Similarly, many studies [4, 18, 30] have stated that CRP and LYM are important biomarkers in determining the presence and severity of the disease. It also increased the advantage of this decision tree (Fig 3) , in identifying patients (Table 8) with higher accuracy than direct blood values and individual achievements of all biomarkers (Table 5 ). Here Table 8 In this study, biomarkers (derived) that were effective in distinguishing ICU patients from non-ICU patients to determine disease progression were found to be more successful than the diagnostic performance of direct blood values (Table 7) A decision tree was drawn that interprets blood values and biomarkers that are effective in classifying patients into ICU and non-ICU together with a simple decision (Fig 4) . In determining patients to ICU or non-ICU, the total classification accuracy rate of the decision tree, which was created with d-CIT and d-PPT biomarkers in a statistically significant way, was 93.5 %, the accuracy rate of classifying ICU patients was 26.8 %, and the accuracy rate for classifying non-ICU patients was 98.1 % (Table 9 ). It was not desirable that the general accuracy rate of the decision tree obtained in Here Table 9 However, the success of this decision tree (Fig 4) in identifying non-ICU patients was much higher than the individual success of the predictors presented in Table 7 for classifying non-ICUs. In addition, the d-CIT and d-PPT biomarkers, which are statistically significant in Fig 4, perfectly classify non-ICU patients and identified a high percentage of ICU (Table 7) . Table 7 should be considered. However The presence of CRP-derived biomarkers at the root of the tree provides a confirmation of our interpretable Decision Tree (Fig 4) of the robustness of the approach to determining disease severity.Similarly, [31] and [32] reported that CRP and ESR increased with disease severity. The biomarkers previously (LCR, NLR, d-NLR, PLR) used to determine the severity of the disease were lower than the biomarkers derived from the study in determining ICU (Table 7) [ In another study, it was noted that the platelet count decreased in the ICU group and was associated with the severity of the disease [34, 37, 38] . Similarly, in this study, the platelet count was similar between non-ICU and ICU and was greatly reduced compared to the control group. D-Dimer was found to be high in most of both ICU and non-ICU patients. Changes in routine laboratory data in the study were similarly seen in other studies [19, 23, 26, 34, 39, 40] . In some studies, the difference in prametre values is mostly in the same direction in this study [9, 27] . Among the reasons for the difference, the severity stage of the disease, sampling time, additional diagnosis of the patients and the number of patients included in the studies can be said. Various independent biomarkers used in the study need to be tested in the diagnosis and prognosis of many other infectious diseases. The low number of ICU patient groups compared to the non-ICU group was one of the limitations of this study. In addition, since it is a retrospective study from the records, inability to access comorbidity data and being singlecentered are other limitations of this study. Retrospective studies naturally lack control of variables; Therefore, prospective cohorts are also needed to validate our study data. In addition, the data used in this study were obtained from the Covid-19 patient population in the spring, summer and autumn seasons. Therefore, some parameter values in this study showed seasonal or periodic differences from similar studies. Multicenter studies in larger patient groups will further clarify the importance of routine laboratory parameters in Covid-19. Authors state no conflict of interest. Informed consent was obtained from all individuals included in this study. The study protocol was approved by the Institutional Ethics Table 1 Blood routine parameter and biomarker values of non-ICU group and Control group on admission. Table 9 . Estimated accuracy of blood parameters and independent biomarkers in the differentiation of non-ICU and ICU (ie, prognosis of the disease) using CHAID analysis. Abnormalities of peripheral blood system in patients with COVID-19 in Rapid and accurate identification of COVID-19 infection through machine learning based on clinical available blood test results Diagnostic utility of clinical laboratory data determinations for patients with the severe COVID-19 Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads and lung injury Biomarkers associated with COVID-19 disease progression Hematologic predictors of mortality in hospitalized patients with COVID-19: a comparative study Which lessons shall we learn from the 2019 novel coronavirus outbreak? Receptor recognition by the novel coronavirus from Wuhan: an analysis based on decade-long structural studies of SARS coronavirus COVID-19: Pathophysiology, treatment options, nanotechnology approaches, and research agenda to combating the SARS-CoV2 pandemic COVID-19: a promising cure for the global panic COVID-19): experience in Zhejiang Province Clinical characteristics of refractory COVID-19 pneumonia in Neutrophil-to-lymphocyte ratio as an independent risk factor for mortality in hospitalized patients with COVID-19 The diagnostic and predictive role of NLR, d-NLR and PLR in COVID-19 patients Laboratory predictors of death from coronavirus disease 2019 (COVID-19) in the area of Valcamonica Could ferritin level be an indicator of COVID-19 disease mortality? Thrombocytopenia is associated with severe coronavirus disease 2019 (COVID-19) infections: a metaanalysis The hyperferritinemic syndrome: macrophage activation syndrome, Still's disease, septic shock and catastrophic antiphospholipid syndrome Clinical characteristics of coronavirus disease 2019 in China Hematologic parametersin patients with COVID-19 infection Lymphopenia predicts disease severity of COVID-19: a descriptive and predictive study Molecular, serological, and biochemical diagnosis and monitoring of COVID-19: IFCC taskforce evaluation of the latest evidence The laboratory tests and host immunity of COVID-19 patients with different severity of illness High expression of ACE2 receptor of 2019-nCoV on the epithelial cells of oral mucosa IL-19 induces production of IL-6 and TNFalpha and results in cell apoptosis through TNF-alpha The significance of neutrophil-lymphocyte ratio, plateletlymphocyte ratio and lymphocyte-monocyte ratio in predicting peripheral arterial disease, peripheral neuropathy, osteomyelitis and amputation in diabetic foot infection Association of inflammatory markers with the severity of COVID-19: a meta-analysis Clinical characteristics of 140 patients infected with SARS-CoV-2 in Wuhan, China Fasting blood glucose at admission is an independent predictor for 28-day mortality in patients with COVID-19 without previous diagnosis of diabetes: a multi-centre retrospective study Hemostatic laboratory derangements in COVID-19 with a focus on platelet count Abnormal coagulation parameters are associated with poor prognosis in patients with novel coronavirus pneumonia Understanding the COVID-19 coagulopathy spectrum Prediction of severe illness due to COVID-19 based on an analysis of initial fibrinogen to albumin ratio and platelet count The association between severe COVID-19 and low platelet count: evidence from 31 observational studies involving 7613 participants Hematologic, biochemical and immune biomarker abnormalities associated with severe illness and mortality in coronavirus disease 2019 (COVID-19): a meta-analysis Laboratory abnormalities in children with mild and severe coronavirus disease 2019 (COVID-19): a pooled analysis and review ESR: erythrocyte sedimentation rate; INR: international normalized ratio PT: prothrombin time; LCR: lymphocyte to C-reactive protein ratio; d-CL: CRP to lymphocyte ratio; d-CI: CRP times INR; d-CT: CRP times Troponin; d-TI: Troponin times INR; d-PPT: PT times Procalcitonin times Troponin; d-CIT: CRP times INR times Troponin (see method section) (see method section